Discovery

Inspiration

We have been very interested in how to optimize learning, and have had ideas before to increase the rate of learning. The L3Harris challenge to inspire and teach <12 year olds was the perfect avenue to build out this idea.

What it does

It takes a concept the student wants to learn, and converts it to a topological map of concepts in the order they need to be learned in using LLMs. Then we combine that with state of the art spaced repetition techniques, which outperform standard Anki techniques.

How we built it

Leveraging the power of Large Language Models (LLMs), we developed an algorithm capable of understanding and structuring any given STEM concept into a digestible, topological map. This map outlines the essential topics and their interconnections, creating a personalized learning pathway for each user. To implement the spaced repetition system, we studied cognitive science principles, integrating algorithms that adapt the timing of concept reviews to the user's learning pace, enhancing memory retention. Our tech stack includes Python for backend development, React for the user interface, and cloud services for hosting our application and managing data.

Challenges we ran into

One of the main challenges was ensuring the LLMs could accurately interpret and break down complex STEM concepts into simpler, understandable segments for children under 12. Balancing technical accuracy with child-friendly language required iterative testing and refinement. Another significant challenge was designing an effective spaced repetition algorithm that could dynamically adjust to each user's learning curve. Additionally, we faced technical hurdles in integrating various software components into a seamless user experience and ensuring our platform was scalable and secure.

Accomplishments that we're proud of

We are incredibly proud of creating a platform that truly customizes the learning experience for young STEM enthusiasts. "Discovery" not only simplifies complex subjects but also makes learning interactive and fun. Seeing our initial prototypes come to life and receiving positive feedback from early testers has been immensely rewarding. Moreover, our ability to overcome technical challenges and work collaboratively as a team, leveraging each member's unique skills and perspectives, has been a significant accomplishment in itself.

What we learned

Throughout this project, we've gained profound insights into the intricacies of educational technology and the specific needs of young learners. We've learned about the importance of user-centered design, especially when creating educational content for children. The technical challenges pushed us to deepen our understanding of AI, machine learning, and software development practices. On a broader level, this project has taught us about resilience, teamwork, and the transformative power of technology in education.

What's next for Discovery

Looking ahead, we plan to expand "Discovery" by incorporating more subjects, languages, and personalized learning features. We aim to collaborate with educators and cognitive scientists to refine our algorithms and content further. Additionally, we're exploring partnerships with schools and educational organizations to reach more students.